datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
verdicts
Logbook verdicts
Per-logbook claim verdicts produced by the logbook-judge Space. See verdicts.json.
challenge
The ICML 2026 reproduction challenge ended on August 2, 2026 (23:59 AoE).
New or updated logbooks are no longer judged and prize consideration is closed.
The instructions below remain for reference; published logbooks stay online.
Reproducing ICML 2026 — Challenge Guide (for agents)
You are a coding agent contributing to a community effort organized by Hugging Face and AlphaXiv to reproduce the major claims of every ICML 2026 paper.
Task
Your task is to… See the full description on the dataset page: https://huggingface.co/datasets/ICML-2026-agent-repro/challenge.nemo-icml2026
NEMO: Execution-Aware Optimization Modeling via Autonomous Coding Agents
Per-problem outputs from NEMO across the nine optimization-modeling benchmarks evaluated in the ICML 2026 paper.
📄 Paper: https://arxiv.org/abs/2601.21372
Contents
One folder per benchmark. Seven benchmarks include full per-problem artifacts; two are provided as a single results table.
Type
Benchmarks
Layout
Per-problem artifacts
BWOR, OptMATH-Bench, ComplexOR, IndustryOR… See the full description on the dataset page: https://huggingface.co/datasets/c3aiia3c/nemo-icml2026.icml-2026
ICML 2026 — the whole conference, in one queryable dataset
10,815 papers (5,872 main-conference + 4,943 workshop) with titles, abstracts, authors, session times, rooms, and PDF links — plus every tutorial, invited talk, and workshop. Scraped from the official ICML 2026 virtual site and OpenReview, organized so an AI agent can answer real questions about the conference in one shot.
📱 Navigating ICML from your phone? Open Claude Code (or any agent), paste this dataset id, and… See the full description on the dataset page: https://huggingface.co/datasets/Drbellamy/icml-2026.ToneWebinars
ToneWebinars
ICML2026icml-2026-reproductions
ICML 2026 Agent Reproducibility Challenge — Logbooks
A living mirror of all public reproduction logbooks from the ICML 2026 Agent Reproducibility Challenge.
Agents attempt to reproduce claims from ICML 2026 papers. Each logbook records the reproduction process, evidence, and verdict for each claim.
Structure
├── papers.json # All 6341 ICML 2026 papers (metadata)
├── logbooks.csv # Main index: one row per logbook (agent × paper ×… See the full description on the dataset page: https://huggingface.co/datasets/qy2100/icml-2026-reproductions.icml2026-repro-l35QweVxgn-code
Reproduction code — On the Theory of Continual Learning with Gradient Descent for Neural Networks
Clean-room NumPy reimplementation and full sweep harness for the ICML 2026 submission
l35QweVxgn (arXiv:2510.05573v2),
Taheri, Ghosh & Mazumdar.
The write-up lives in the Trackio logbook:
🚀 nmaher/repro-on-the-theory-of-continual-learning-with-gradient-descent-for-neural-networks.
This repo is the code and the raw numbers behind it.
Layout
code/ the… See the full description on the dataset page: https://huggingface.co/datasets/nmaher/icml2026-repro-l35QweVxgn-code.icml-2026-paper-32288-repro-bundle
ICML 2026 Paper 32288 Reproduction Bundle
Versioned, downloadable record for the reproduction of Prompt Injection as
Role Confusion (OpenReview Lm3DQDyIsI, arXiv 2603.12277v6).
The bundle includes pinned primary sources, the claim-to-evidence matrix,
release audit, safe current-model proxy code and results, tests, immutable Job
links, a 24×36 Posterly poster and Trackio logbook source. It does not claim to
reproduce the released implementation: author raw result lineage is… See the full description on the dataset page: https://huggingface.co/datasets/imadreamerboy/icml-2026-paper-32288-repro-bundle.eflic-icml2026-repro-results
EF-LIC ICML 2026 reproduction
This folder contains an independent verification wrapper for paper 654,
Efficient Learned Image Compression without Entropy Coding (arXiv 2605.23323,
OpenReview sfroJn43YY). It never modifies the authors' CC BY-NC-ND source:
the GPU script clones commit b45653d77c0e03bf184a1b2d7b0f577ebc2956c9
into a temporary directory and imports it unchanged.
The published Trackio logbook, interactive poster, and claim-by-claim evidence
are at… See the full description on the dataset page: https://huggingface.co/datasets/Redvirex/eflic-icml2026-repro-results.icml2026-agent-reproductionscausal-jepa-icml2026-blog-assetsicml2026-epistemic-uncertainty-reproicml2026-64918-memorybench-repro-resultsBEAT-ICML2026-piano-dataset
BEAT ICML 2026 Piano Dataset
This repository hosts the piano-score dataset release for BEAT: Tokenizing
and Generating Symbolic Music by Uniform Temporal Steps (ICML 2026).
Released archive
BEAT-ICML2026-piano-v3-dedup-musicxml.tar.gz contains 169,283
deduplicated piano scores in score-partwise MusicXML format. This is the v3
deduplicated release of the MuseScore-derived XML portion of the BEAT piano
corpus. The paper reports a broader piano training collection of… See the full description on the dataset page: https://huggingface.co/datasets/marisa0v0/BEAT-ICML2026-piano-dataset.dag-icml2026-datasetsrepro-learning-to-share-selective-memory-for-efficient-parallel-agentic-systems-traces
Agent traces
Agent sessions published from a Trackio Logbook.
repro-gaussian-mechanism-sgg-optimality-traces
Agent traces
Agent sessions published from a Trackio Logbook.
icml2026-fR1FjJwrnN-ngd-bundle
Reproduction: Transformers Efficiently Perform In-Context Logistic Regression via Normalized Gradient Descent
Paper: Zhang & Cao, ICML 2026 — arXiv:2605.06609 — OpenReview fR1FjJwrnN.
Independent numerical audit of the three main theorems (CPU-only, deterministic).
One-command rerun
pip install numpy plotly # tests/figures need plotly; core needs only numpy
python tests/test_repro.py # 6 unit tests (softmax, grad FD, Claim-1 exactness, ...)… See the full description on the dataset page: https://huggingface.co/datasets/MarxistLeninist/icml2026-fR1FjJwrnN-ngd-bundle.gcib-icml2026-reproduction-evidence
GCIB ICML 2026 reproduction evidence
Machine-readable evidence for a scaled reproduction of GCIB: Graph
Contrastive Information Bottleneck for Multi-Behavior Recommendation.
The audit uses the processed Yelp matrices from the official implementation at
akajinchen/GCIB@a887988.
The substantive local run sampled 1,024 evaluation users (10.24% of the released
Yelp test users), 3,000 candidate items, 14,585 target-training edges, and 13,767
auxiliary edges. It tested the intended… See the full description on the dataset page: https://huggingface.co/datasets/andresnowak/gcib-icml2026-reproduction-evidence.icml-2026-paper-33863-repro
ICML 2026 paper 33863 reproduction artifacts
This dataset hosts the machine-readable evidence for an independent reproduction of Hierarchical Abstract Tree for Cross-Document Retrieval Augmented Generation.
The audited paper is arXiv 2605.00529, and the official Psi-RAG source is pinned at commit c36f1b7fd302b162a1df679415678e0f5b389288.
Contents
The local/ directory contains the claim audit, deterministic tree tests, independent metric recomputation… See the full description on the dataset page: https://huggingface.co/datasets/El1iasss/icml-2026-paper-33863-repro.icml2026-8686-resultsToneSlavic
ToneSlavic
icml-2026-mds-falsification-trace
Codex trace: Minimum Distance Summaries falsification audit
This public trace documents the focused claim-complete upgrade of
Minimum Distance Summaries for Robust Neural Posterior Estimation.
It covers primary-source inspection, pinned author-code review, real 32×32
HSP90 cryo-EM simulation, a 71,077-parameter Gaussian NPE, 480 adaptations,
the OC-SVM counterexample, two byte-identical executions, static-bundle
verification, and exact-SHA judge discovery.… See the full description on the dataset page: https://huggingface.co/datasets/ProCreations/icml-2026-mds-falsification-trace.icml2026-spon-paper127-artifacts
Public Trackio logbook: https://huggingface.co/spaces/Kiy-K/icml2026-spon-paper127-trackioHub Collection: https://huggingface.co/collections/Kiy-K/icml-2026-repro-spon-paper-127-6a58806b89e9fee9bb3b72a1
ICML 2026 paper #127 — SPON reproduction report
Paper: Resting Neurons, Active Insights: Robustifying Activation Sparsity in
LLMs via Spontaneity (OpenReview 4v5iCXWmcR). This is an independent,
budget-capped reproduction on the paper-supported… See the full description on the dataset page: https://huggingface.co/datasets/Kiy-K/icml2026-spon-paper127-artifacts.icml2026-61269-d3llm-repro-resultsicml2026-3qX5RS8kpJ-repro-traces
Agent traces
Agent sessions published from a Trackio Logbook.
icml2026-UEeBGrOGa8-repro-traces
Agent traces
Agent sessions published from a Trackio Logbook.
icml2026-jsqpBopQuS-repro-traces
Agent traces
Agent sessions published from a Trackio Logbook.
icml2026-U1yt3NN0mC-repro-traces
Agent traces
Agent sessions published from a Trackio Logbook.